|2021||Feb Mar Apr May Jun Jul|
|2020||Jan May Jun Jul Aug Sep Oct Nov Dec|
|2019||Jan Feb Mar Apr May Jun Jul Aug Oct Nov|
|2018||Jan Feb Mar Apr May Jun Jul Aug Oct Nov Dec|
|2017||Jan Feb Mar Apr May Jun Jul Aug Oct Nov Dec|
Event(s) on January 2012
- Tuesday, 3rd January, 2012
Title: CMIV Colloquium: Large Scale Ice Sheet Modeling and Simulation Speaker: Professor Esmond G. Ng, Lawrence Berkeley National Laboratory, USA Time/Place: 11:30 - 12:30
FSC1217, Fong Shu Chuen Library, HSH Campus, Hong Kong Baptist University
Abstract: Understanding the changing behavior of land ice sheets is essential for accurate projection of sea-level change. The dynamics of ice sheets span a wide range of scales. Localized regions such as grounding lines and ice streams require extremely fine (better than 1 km) resolution to correctly capture the dynamics. Resolving such features using a uniform computational mesh would be prohibitively expensive. Conversely, there are large regions where such fine resolution is unnecessary and would represent a waste of computational resources. This makes ice sheets a prime candidate for adaptive mesh refinement (AMR), in which finer spatial resolution is added where needed, enabling the efficient use of computing resources. The Berkeley ISICLES (BISICLES) project is a collaboration among the Lawrence Berkeley National Laboratory, Los Alamos National Laboratory, and the University of Bristol in the U.K. We are constructing a high-performance scalable AMR ice sheet model using the Chombo parallel AMR framework. The placement of refined meshes can easily adapt dynamically to follow the changing and evolving features of the ice sheets. We also use the vertically-integrated treatment of the momentum equation due to Hindmarsh and Schoof (2010), which permits additional computational efficiency. Autotuning techniques are being deployed to improve performance of key computational kernels. Linking to the existing Glimmer-CISM community ice sheet model as an alternative dynamical core allows use of many features of the existing model, including a coupler to CESM. We present preliminary results showing the effectiveness of our approach, both for simple benchmark problems which validate our approach, and for application to regional and continental-scale ice-sheet modeling.
- Thursday, 12th January, 2012
Title: DLS: High-Dimensional Statistical Inference: From Vectors to Matrices Speaker: Prof. T. Tony Cai, The Wharton School and University of Pennsylvania, USA Time/Place: 15:00 - 16:00 (Preceded by Reception at 14:30pm)
SCT909, Cha Chi-ming Science Tower, HSH Campus, Hong Kong Baptist University
Abstract: Driven by a wide range of applications, high-dimensional statistical inference has seen significant developments over the last few years. These and other related problems have also attracted much interest in a number of fields including applied mathematics, engineering, and statistics. In this talk I will discuss some recent advances on several problems in high-dimensional inference including compressed sensing, low-rank matrix recovery, and estimation of large covariance matrices. The connections as well as differences among these problems will be also discussed.